The missing shock: the macroeconomic impact of British Privatizations

Abstract

In this paper the impact of privatization on macroeconomic performance in the United Kingdom is tested using quarterly data from 1979 to 1999. Privatization proceeds have been included in a simple analytical framework dealing with both demand and supply-side of the economy. Multivariate cointegration techniques have been used in order to consider the nonstationarity of the time series involved. The empirical results show that privatizations have no long-run effects on output in the UK. This result is consistent with microeconomic evidence that shows that in the UK ownership change per se had little impact on long term productivity trends. Moreover it is found that privatization proceeds have contributed to sustaining public expenditures.

Acknowledgments

M. Florio is grateful to the Ministery of University (MIUR) for financial support (40% cofinancing Head of Unit G. Bognetti). The authors are also very grateful for comments on previous drafts to D. Checchi, M. Galeotti, A. Missale, M. Sawyer, J. Leape, D. Newbery and other participants in the First Milan European Economy Workshop, June 2002, and to a seminar in the Department of Economics, University of Milan, March 2003. The usual disclaimer applies.

Notes

This section draws from Florio (2004).

Florio and Manzoni (2004) show that in the UK this may have been around 20% in 24 hours for several initial public offerings.

Incidentally, it may be noted that with underpricing and higher mark-up there is a regressive redistribution of wealth from the taxpayers and consumers in favour of the shareholders. This may bring down consumption and output in the long term (particularly if some of the shareholders are foreign investors).

By comparison, Wallis et al. (1987) and Florio (1990) review simulations of macroeconomic impacts of seven econometric models of the UK economy in the 1980s and find:

	1.	3.25% increase of public expenditures financed by high-powered money increase GDP by 0.6–0.9% for most models.
	2.	5% decrease of marginal rate of income tax (e.g. from 30% to 28.5%) has a typical impact around 0.2% or GDP
	3.	10% decrease of VAT rate (from 15% to 13.6%) has an impact around 0.2% on GDP.

These examples have a dimension similar or inferior to the yearly flow of privatization proceeds on most years, and it can be assumed that privatization proceeds variable is not negligible in public finance terms.

OECD (1998, fig. 11, p. 55).

Oulton (1995) and Brown et al. (1997).

Cf. OECD Table A, p. 176.

See Newbery (2000). Appendix, chapter 1.

The OECD data on the other hand are not in full agreement with the picture that emerges from HM Treasury (1999, p. 131) which, at 1995 prices, shows growth in the ratio of non-residential investments of GDP between 1982 and 1989, then drop off until 1994 and subsequently a noticeable recovery.

The inflationary impact of cuts in capital account public spending was studied by Buiter (1990), in a different context. He shows, applying a model of a small open economy, that public investments cuts have four separate effects:

	1.	A direct effect on spending, which presumes a reduction of the budget deficit of exactly the same size as the investment cut;
	2.	A direct effect on budget revenues, equal to the variation in cash flows from the management of public capital;
	3.	An indirect effect on tax revenue, as a consequence of the influence of lower investments on aggregate demand and thus on fiscal receipts;
	4.	Lastly an effect on the demand for money which, given a certain rate of real interest and of inflation, operates through the income (or wealth) of the agents.

Whilst it is clear, also intuitively, that ceteris paribus effect (a) is deflationary and that effect (b) is ambiguous, the last two effects are unambiguously inflationary. It is therefore possible that the two groups of effects balance each other out.

In order to determine the permanent effect of privatization on GDP growth, the author estimates equations of the following form: y _it  = μ + αy _{i ,} t ₋₁ + δp _it  + u _it where y _it is the GDP growth, p _it are privatization proceeds expressed as a share of GDP and u _it are the residuals.

Univariate cointegration techniques, however, have been applied in a first draft of the paper, see Florio and Grasseni (2003).

The data are quarterly and cover the period 1972 : 2–1999 : 1. Data are at constant 1995 prices and seasonally adjusted; data at current prices have been deflated by the GDP deflator. The main sources for the data is ONS except for UNEMPL data that are from OECD. Variables: p _t : log of Privatization Proceeds (The main source of Privatization Proceeds are ONS Financial Statistic Explanatory Handbook, 2000 edition, and HM Treasury Public Expenditure: Statistical Analyses.; y _t : log of Gross Domestic Product; g _t : log of Total Current Government Expenditure; inv _t : log of Gross Fixed Capital Formation; i _t : Three Month Yield, Treasury Bills; π _t : RPI: Percentage change over 12 months – all items; u _t : Unemployed percentage of total labour force (source: OECD).

The seven largest unit roots are (0.95, 0.90 ± 0.16i, 0.82, 0.55 ± 0.06i, 0.39).

The test procedure which test restriction on a single vector, leaving the other vectors unrestricted (Johansen and Juselius, 1992), suggests that the output relation cannot be rejected by a χ²(1) = 0.0005, with a p-value = 0.98.

Estimation results reported in Table 2 highlight that none of the variables are weakly exogenous for the cointegrating relations.